A little background, first:
I've always loved doing time-lapses of weather. Storms in coastal NC, snow in Fort Drum, NY, clouds in general, etc. I now live in a holler in eastern KY, and my webcam is set up in my window looking toward the NNW. Until a little over a month ago, I would only run a time-lapse every once in a while. Then I started noticing that my holler is a pretty awesome microscale area (we also get a ridge/valley split a lot of calm nights). This brings me to the time-lapses.
The first video is from June 25. I set the camera up before a few showers passed through the region and left it running...thinking maybe I would catch some building clouds and then a little rain. When I reviewed the video, instead what I caught was a shower that inexplicably started rotating! Mind you, this is not a big storm...there was no real shear, no front...it was just a little pop up shower. At first I thought I was seeing things, but when I looked at the velocity on the radar, there were indeed a couple of red pixels next to a bunch of green pixels! The beam on the 0.5 degree slice looks about 1,800 ft above the ground. So, I hypothesized that what happened is the shower just happened to be coming in the right direction for our holler to cause the inflow to "rotate". It went down the length of the holler, and as it passed over the curved section, it started rotating...then once the updraft passed that area, the rotation dramatically decreased. I seen clouds going up the sides of the mountains quite a bit...as the warm, moist air is forced up the side, it condenses into scud and keeps working it's way up the mountain. What I had not seen, however, is this "forcing" actually spin...let alone spin enough to reach up to the base of the cloud and then cause the cloud to start rotating! I was kind of skeptical at first, but could not think of any other plausible explanation. One of my friends at the weather office in Jackson said he believed that the terrain influenced the rotation as well, and was quite surprised the rotation was enough to be picked up on radar! Another one of the forecasters (who was on duty at the time) was surprised at how much the shower was rotating in the video, as well. The reason I think this particular storm rotated so well, compared to the two others I've captured, is that it was moving slower over the area...and it didn't just glance by, it actually came almost straight down the holler.
Here is the short video (focused on just the rotation), followed by the full time-lapse:
http://youtu.be/jac6t0j3RO0
http://youtu.be/hHxLcBfn9u4
Please note:
These time-lapses I am posting are all taken the same. First off is frame capture...I have my webcam set up to capture a frame every 5 seconds. Then, the playback speed is about 20 frames a second. If you do the math...that comes out to about 1.7 minutes of "actual" time covered in 1 second of video...the whole video covers about 26 minutes of time (for the short, rotation focused only one).
The second storm I captured rotation on was on July 24, and you could only see rotation at cloud level as well. This storm came in at a different angle, but hopefully (if youtube does not kill the quality) you should be able to see some scud twirling around each other in about the same spot the shower in the first video started rotating. If you need to, go to settings on the video itself, and select the 720p HD option...it'll "crisp" the video up some.
http://youtu.be/Ko4VUPXINt8
The third storm I caught this afternoon (July 27), and it showed me exactly what I hoped to see! Today was a really humid, rainy day...which caused scud to form darn near all the way to the ground. At first, as the updraft region approaches, you can see the scud being pulled into the updraft from right to left (storm motion was left to right). Then suddenly, as the updraft region crosses into the holler, the scud start rotating. Now...today had quite a bit of shear, but it did not appear that the base of the cloud was rotating beforehand, and the scud started rotating in almost the same spot as the other two videos. The only difference, this time, is that with the scud...you could see the whole column rotating from the ground to the cloud base! Keep in mind that this is a time lapse (1 frame every 5 seconds, 20 frames a second playback). The whole length of the rotation is about 3 seconds or so on the video...which would cover approximately 5 minutes of real time. No, this is not a tornado (it is not a violently rotating column of air)...it isn't rotating that fast, the clip being sped up makes it appear that way. And again, please switch to 720p HD in the settings and then watch the video!
http://youtu.be/_PqyjDxH6G8
As a bonus for making it this far...here's some cool clouds, followed by heavy rain from a storm on June 20 (I believe this was an attempt to form a wall cloud, since the cloud is sloping toward the precip...it also appears to be mountain caused). This time-lapse had a capture rate of 1 frame every 10 seconds:
http://youtu.be/pGD31Jpwn14
Here is the radar image showing the rotation from June 25...the timestamp on the velocity product was 21:30Z (that's 17:30/5:30 pm EDT). That means that scan was taken right at the end of the time-lapse clip. The 5:25 pm EDT scan didn't have any rotation on it...but I do not have a copy of that image.
